ad-76 7 2 99 magnetic bubble materials jerry w. moody, … · this is the final report on the...
TRANSCRIPT
,l"«1 ' i' "' ■""■•"'"iiwiii ■■>■'•-•■ 1*1. iwpp«»» in i . i ■■■■ «u nun i.imiMMpnanjpiiiiiiian mmmmmmmrimmmimm
AD-76 7 2 99
MAGNETIC BUBBLE MATERIALS
Jerry W. Moody, et al
Monsanto Research Corporation
Prepared for:
Advanced Research Projects Agency
II August 1973
DISTRIBUTED BY:
im National Technical Informatien Service U. S. DEPARTMENT OF COMMERCE 5285 Port Royal Road, Springfield Va. 22151
.H'.1>^^!'^,.1.. -i.-.,..-.... ■ .. ..^.X.-^.: IL^U^V^ilt, --^.^l^irtniliMia^-^---'t-1 ^■-■-—--^■-"^|-||i|- iV J.|| -
FMiiisiiiii ■.,>iip»-i<iiw> K.III» L'^oj..-"" < ü i jain«! »wi J •'■ "" JII. H™U«I«.IPIP i. ■ijiiiniin ■!■ m n i n ma m u.! i>< llini>iili|>!U)um
MRC-SL-386
FINAL REPORT
CD |s«
MAGNETIC BUBBLE MATERIALS
Monsanto Research Corporation St. Louis, Missouri 63166
Jerry W. Moody, Robert M. Sandfort, Roger W. Shaw
Michael C. Wlllson, Joseph T. Cheng, and Richard J. Janowiecki
Contract No. DAAH01-72-C-1098
Distribution of this document is unlimited
ARPA Support Office Research, Development, Engineering,
and Missile Systems Laboratory U.S. Army Missile Command
Redstone Arsenal, Alabama is
\ v
.
Sponsored by:
Advanced Research Projects Agency ARPA Order Nr. 1999
Reproduced by
■■iL.^,,....^.-x.-..J...'..-I^-^VL.^..■...-.,.,:■ ■.,.■ ,-..i.v-,,.w-. .^i.^..^i^.^^l^.^'T|t[lfjitillirflfil'ffitrVitril'iifl^liliMlrMtlrtiJlilVlirf-ttTlii/iili-aTi'
NATIONAL TECHNICAL INFORMATION SERVICE
U S D.nit.it nenl of Commefce SpringlioldVA 22151
aaaa mm« abfetfa ■..■-IIII..-..I- .<■.„. a—■ •■■i-m. „„^
n
^.^t..J.^.',.^. .^.^.. ■■.,.,■■...., . ,„.;.... -o-^-.»^.-.^.^'^.^..^^ .^.. ..-.^.--vj^.j.riLj^t.i..
"^■-'^TTW^^wiair*WTOf'»^^
UNCLASSIFIED Sfcunty Classification
DOCUMENT CONTROL DATA -R&D (Secr.ly fl,*„l,c„ion o/ r)(l«. body t.> afc.lr«.-» W MrMhtj mmolatlon mu.s, hf -nlered when (he overall ,.p,„f U rlasilfled)
Monsanto Research Corporation 800 North Lindbergh Boulevard St. Louis, Missouri 6 3166
3 HEPOUT TITLE
I». REPORT SECURITY CLASSIFICATION
UNCLASSIFIED 2b. GROUP
MAGNETIC BUBBLE MATERIALS
< DESCRIPTIVE NOTES (Type ol report and Inclusive dales)
Final Report, 11 January-11 July 1973 S A'j TMORIS» (First nttme, middle initial, last name) — ——_
J. W. Moody, Robert M. Sandfort, Roger W. Shaw, et al.
6 RtPOW T D A T E
20 September 1973 Bo CONTRACT OH GRANT NO
DAAHO1-72-C-1098 6. PROJEC T NO
ARPA Order No. 1999
7a. TOTAL NQ. OF PAGES
INATOR'S REPORT NUMBER(S)
76. NO OF REFS
28
MRC-SL-386
ab. OTHER REt : RT NO(SI (Any other numbers that may be assigned in t s nport}
10 OIS TRIBUTION STATEMENT
Distribution of this document unlimited
11 SUPPLEMENTARY NOTES
13 ABSTRACT
12- SPONSORING MILITARY ACTIVITY
Advanced Research Project Agency Washington, D.C.
The Czochralski growth of defect-free, non-magnetic garnet crystals and the preparation of epitaxial magnetic bubble garnet films by liquid phase tpitaxy (LPE) and arc-plasma spraying (APS) were investigated. Two garnet systems, Eu- Yb-Y and Sm-Y, were identified as being of particular practical interest. These garnets exhibit reasonable mobilities, satisfactory quality factors, and good temperature stability. Device-quality films can be deposited by LPE from PbO-Bj O3 solution on (111) GGG substrates. Methods were developed for growing epitaxial garnet films by APS. However, the films were usually contaminated with crystallites of orthoferrite, hematite, magnetoplumbite or other phases. Low-defect-density films were not pre- pared and the elimination of unwanted phases by compositional control appears to be a formidable problem. All recent developments indicate that the LPE "dipping" growth of garnet films is a commercially viable process. Core-free single crystals of GGG with defect densities of less than 5 per cm2 can be grown readily by the Czochralski method. Such crystals, up to 38 mm in diameter, are now available commercially.
DD .Fr„1473 UNCLASSIFIED
K Security Classification
-■"■"— '- mSM Mttfll ■ ■ ...-.^..^i^mJ^.J^imaUMIJ^ll^^^^ ■,v........„,.v...i:.,,. :.- ,^,.,JL..,.> , j,..,...;.- „-..,. ■> ■■' ■—- ^-^ - uu,*,..
vttmm&T&***™v!?*??™T*T*™r^^
UNCLASSIFIED Secimty Ctaaaification
'.«V WONOS
Magnetic bubble materials
Magnetic garnets
Gadolinium gallium garnets
Crystal growth
Liquid phase epitaxy-
Chemical vapor deposition
Arc plasma spray-
Substrate preparation
Rare-earth iron garnets
NOLC
LINK B
N04.B
LINK C
NOLK
1 UNCLASSIFIED
Security CUasificction
-•■ ^ -•"'-'—■"""■""—niirimiMiii - mtm ■ , - ■ ,t*t
?^—'^r-T«^^^v.1^.'M■<■|!"-wr?**-^.**''**r-**-*'W'■vj«*u..«wru^rFwr.'.1 iw.ffP^Hf»^^wv^'r-t^w^J^WV\yr»T"^™*?»?^^.-^ii«/-»'w«/.v^«^*»w»jii)rji»!_^^!M^fiy^.wii|wi^!»-jv?"r-i-4*T«(W^PT-^I 'WMW.T^-I!wi!»iJ.■ i,J". 1.1».*«nuiai»i.!■ ,1.^•,».i.-.....w,,,mm>.*yK*fmi">j..., ,\..mmsm
FINAL REPORT
MAGNETIC BUBBLE MATERIALS
Monsanto Research Corporation St. Louis, Missouri 63166
Jerry W. Moody, Robert M. Sandfort, Roger W. Shaw, Michael C. Willson, Joseph T. Cheng, and Richard J. Janowiecki
August 11, 1973
Contract No. DAAH01-72-C-1098 ■ |
'3 :973-
.- Distribution of this document is unlimited J
ARPA Support Office Research, Development, Engineering
and Missile Systems Laboratory U.S. Army Missile Command Redstone Arsenal, Alabama
Sponsored by: Advanced Research Projects Agency
ARPA Order No. 1999
IC
■^^..^..■-■.■...^■.^.^^^ *^^^..^^^...,.^ ^1,^i^i..^^:^... .,-.. ,.,,„.„ ,.,.„., miimjg
pp^- ■•-^P^RFri'»prw?f?w=«^^
SUMMARY
This is the final report on the project "Magnetic Bubble Materials,
Phase II, " Contract No. DAAH01-72-C-1098. It covers in detail the work
performed from January 11, 1973 through ,Tuly 11, 1973. The Czochralski
growth of defect-free non-magnetic garnet crystals , the preparation of
epitaxial magnetic bubble garnet films by liquid phase epitaxy (LPE) and
arc-plasma spraying (APS) were investigated. References to earlier work
are included for the sake of completeness.
Two garnet systems, Eu-Yb-Y and Sm-Y were identified as being
of particular practical interest. These garnets exhibit reasonable mobilities,
satisfactory quality factors and good temperature stability. Device quality
films can be deposited by LPE from PbO-BzC^ solution on (11]) GGG sub-
strates.
Methods were developed for growing epitaxial garnet films by APS.
However, the films were usually contaminated with crystallites of ortho-
ferrite, hematite, magnetoplumbite or other phases. Low defect density
films were not prepared and the elimination of unwanted phases by composi-
tional control appears to be a formidable problem.
All recent developments indicate that the LPE "dipping" growth of
garnet films is a commercially viable process. This method is now used
almost exclusively by workers in the field.
Core-free single crystals of GGG with defect densities of less than
5 per cm2 can be grown readily by the Czochralski method. Such crystals,
up to 38 mm in diameter, are now available from commercial vendors.
ib
jj^-.j- .■....,:... ,., ■■.. ..■.;.-...I..-J:-L..^V.-^.^..-,.,.—„^^^-.^■J.-^^-^,^:....-^.-..,;:... -v.--, ^^w^.^-^-^^ir ^.-^^^ -...-. ■ . -.. .■,.■..,. .... L... j -.■ . . U ■■ ^...^ .-..■-^Wr^«..-■..:.■., . ., , .. ...^.
>iiijiui!«iiiwi»n lb«. ,«min.ii«i-«iij wii"«»i 111■ M. .». ii.ii><jjiiiiin^ji(.iigiiij«jj|n9m««q*n!"H«(uj.np<iiHi.win^.<viuvww>i»viIPJIIIIIPWJTII.WHIIPIIJI.IHWKIUJII IUMAIJH jut Mimw1!-! Mtaummm^frnv
TABLE OF CONTENTS
1.0 INTRODUCTION
2.0 CHARACTERIZATION
2, 1 Film Thickness
2. 2 Defect Detection and Location
2. 3 Characteristic Length and Domain Dimensions
2.4 Saturation Magnetization and Magnetic Fields
2.5 Coercivity
2.6 Anisotropy
2.7 Temperature Effects
3. 0 LIQUID PHASE EPITAXY
3. 1 Review
3.2 Experimental
3.3 The Eu-Yb-Y System
3. 3. 1 Introduction
3. 3. 2 Properties of Eu-Yb-Y Garnet Films
3. 3. 2. 1 Magnetization
3.3.2.2 Characteristic Length
3. 3. 2. 3 Anisotropy
3.3.2.4 Coercivity
3.3.2.5 Mobility
Page
1
3
3
3
3
4
4
4
5
7
7
B
9
9
10
10
12
12
15
15
ii
■i^^,.-^...,-.—.-H-.L'V.J..■;—.— ■. ..■■.^.,. . ■A-igm^jum).. -..--—,^..wl^^:^.^.u.1Ji.J„^iJ^^„.ltiii.., V,i;iu.,:. -.U,.v-., ■ ,. ..,:i^:.:=...■ v,.„..!_.■..■. .•.,-.■■...■;■ ■■,.•-,... 1, ■■.,■..
■^n^W^iWfijPTO^m^wiSTn^TW™«^!-»^^^
3.4 The Sm-Y System 21
3.4.1 Introduction 21
3.4.2 Properties of Sm-Y Garnet Films 23
3.4.2.1 Magnetization 23
3.4.2.2 Characteristic Length 23
3.4.2.3 Coercivity 23
3.4.2.4 Anisotropy 26
3.4.2.5 Mobility 2 7
3.4. 2.6 Temperature Coefficient of Characteristic Length 28
3.5 The Nd-Y System 26
3.6 Discussion 30
4.0 PLASMA SPRAYING 31
4. 1 Introduction 31
4. 1. 1 Growth from the Melt 31
4.1.2 Growth from the Solution 32
4. 2 Experimental 33
4.2.1 R.F. Plasma Spraying Station 33
4.2.2 D. C. Plasma Spraying Station 37
4,2. 3 Mate-ial 37
4. 3 Results and Discussion 39
4. 3. 1 D. C. Plasma Spraying 41
4.3.1.1 Epitaxial Crystal Growth from the Melt 41
4. 3. 1. 2 Epitaxial Crystal Growth from Solution 41
111
- ■ - - ■ --■ -^— — —
i ■■■wipi.wviiwiihui. uuiiigumiii mmmm*. ipiii um ■■HiiwiiaMwwwM mmmmmmmmf^""'^1 ^gmmii^^.r'r'^'wmm'm
4. 3. 2 R. F. Plasma Spraying
4. 3. 3 Characteristics of Selected Garnet Films
5. 0 BULK CRYSTAL GROWTH
6. 0 CONCLUSIONS AND RECOMMENDATIONS
7.0 REFERENCES
8.0 APPENDIX
8. 1 Growth Parameters for Eu0i5Yb0jl5Y2.35Fe3#8Galt2O12
8. 2 Growth Parameters for Sm0i36Y2.64Fe3.7Ga1.3O12
43
44
47
48
51
54
54
54
IV
■■•^■- -'■ immr naiininiMli« ■min m i mhiiitfininiTftint iinlr—--"-"-^-- ^ -- -■ ^.—i—^.-L^.^A.^.—a*-^.»—.^.^^i
r-— ..u Ww.v.mj,.m.,.....,. .1. ... , ! .. .._~. iii.iiIIiliW..pi.,jW.a..n>.IVmM1iu ii.awpti . i M ■ ui .ui« i i i IUP^III.»». , 1 ww, mm^fmmwim^mifmmmiiu. mum'» »
LIST OF FIGURES
Page
Figure 1 Magnetization vs Gallium Content for Eu-Yb-Y Films 11
Figure 2 Characteristic Length vs Gallium Content for Eu-Yb-Y Films 13
Figure 3 Optical Hysteresigraph Tracings of Neel Temperature Determination of Eu-Yb-Y Films 14
Figure 4 Optical Response of a Typical Sample of Eu-Yb-Y Garnet to a Step Rise in Bias Field at Various In-Plane Fields 16
Figure 5 Natural Frequency and Decay Time vs In-Plane Field for a Typical Sample of Eu-Yb-Y Garnet 17
Figure 6 Domain Wall Mobility vs In-Plane Field for a Typical Sample of Eu-Yb-Y Garnet 19
Figure 7 Domain Wall Mobility vs Temperature in a Typical Sample of Eu-Yb-Y Garnet 20
Figure 8 Magnetization vs Gallium Content for Sm-Y Garnets 22
Figure 9 Characteristic Length vs Gallium Content for Sm-Y Garnets 24
Figure 10 Coercivity vs Gallium Content for Sm-Y Garnets 25
Figure 11 Mobility vs In-Plane Field for Some SmyY3_yFe5.xGax012
Garnets f 29
Figure 12 Schematic of R.F. Plasma Spraying Station 34
Figure 13 R.F. Plasma Torch Detail 35
Figure 14 D. C. Plasma Spraying Station 38
- - — ^ - - - «tiiiiriiiir-iiiltti'-""",^-"j-'*ni''!'"v'l*''",-''*-'-"!i^'^'"
!•>••-—I."'-- ">"■" — — ' !■• ■•»ip li.j .«n..™«.. ■ ..»iM'i'*iamwimiw^^i*^twmimw\nivm\Mn\mmm\miivmviwm»m&lPmr'™mim^mmw^imr*<if^iigm
Table 1
Table 2
Table 3
Table 4
Table 5
LIST OF TABLES
Uniaxial Anisotropy of Some Sm-Y Garnets
Powder Compositions
Typical Conditions for Growing Epitaxial Garnet Films from a Flux Using D. C Arc Plasma Spraying
Characterization Data of Selected Bubble Films Made by the Method of R.F. Plasma Spraying
Process Conditions Used for Preparing Selected Representative Films by the Method of R.F. Plasma Spraying
Page
26
40
42
45
46
VI
^ ~\,J i.-ii ivrrZiii. .tt«*-^ ttktadtaifeuittU* ::i. ^**diLxaHi*.^..;.X*.-.l-.:-.,:..-. ■. ..-j-:-:, :- .. -,<: :...^.:-...^.:-y.:,^>.^. .:.■:■* .'.L.,i.r.i-.T-)■-■iW.vj-jiit
W*W!W»W^^PW4JWWMW^ nVPWi« FP«U".^J.l'kiJ^IWW»pilI«,M,p».'.^iPfWI'RW^W»W.'W;UiW^U^^?W^fVM^*ffH*HB..™^.»-^J«^Äwr»|«WWWWWU».-".'-^«.■^,-J.«U»,J«.W!^«»fi',.'.^!W'«[U-WflW^Ff-Jf^-'llAffvIJHWJ.^.Ji(!l(^
1.0 INTRODUCTION
Thifj is the final report on the project "Magnetic Bubble Materials -
Ph^se il", Contract No. DAAH01-72-C-1098. This project was a continua-
tion of work performed under Contract No. DAAH01-72-C-0490. The pri-
mary objective of this work was to fulfill a need for a supply of reproducible,
low defect, high quality magnetic bubble materials to support the device
work of various DOD agencies who are investigating the c.pplication of these
materials for end uses of importance to national security. The work has
involved development of methods to grow high quality bulk crystals for use
as substrates, development of substrate polishing and preparation procedures,
studies of epitaxial film growth by three different methods: liquid phase
epitaxy (LPE), chemical vapor deposition (CVD), and arc-plasma spraying
(APS); and development of methods to fully characterize bubble domain
materials.
Two reportsU' ^) have been issued previously which presented in
detail the preparation of garnet films and crystals,and the progress of
bubble materials technology. Two reports' ' 4/ covering characterization
techniques have been issued also. One, '4' a comprehensive, critical survey
of characterization procedures, has been widely accepted as a valuable aid
by workers in the field.
This report covers the work accomplished from. February II, 197 3
through July 11, 1973. Specifically, it discusses the LPE growth of films
of the Eu-Yb-Y and Sm-Y systems and the preparation of garnet films by
APS. In addition, a section is devoted to up-dating characterization
, , ,.,,, ««-««i
procedures. Finally, the report is concluded with a brief survey of bubble
materials technology and recommendations for future work.
>*- — ■-, '■' "■•- ^———■-...>.. ..■ .-. —. wmnnifiMM mrannuififlüMiiiiMniMam linhinn ■ - — ■ *--■■•■^«-«""^•"^■^".■■^ • .
^•0 CHARACTERIZATION
The majority of the methods employed in the characterization of the
garnet films grown under this contract have been dealt with in detail in
earlier reports. (*' 3. 4) The following brief discussion is thus intended
only to summarize these methods and present relevant detail where new
techniques have been instituted since those reports were prepared.
2- 1 Film Thickness
The thickness of the magnetic garnet films has been measured using
optical interference of reflected light(4). Thickness uniformity also has
involved the use of optical interference, in this case in the form of a contour
map caused by interference of monochromatic light reflected from the front
and back of the film. (4)
2. 2 Defect Detection and Location
Defects which impede domain motion are readily observed using a
polarizing microscope and an alternating bias field sufficient to cause most
of the strip domains to move(4). Experience at Monsanto has indicated that
this method, when properly employed, is sufficiently sensitive to detect all
defects likely to cause difficulties in devices of present design.
2. 3 Characteristic Length and Domain Dimensions
Where possible, domain dimensions have been measured on parallel
strip domain patterns in order to improve accuracy. This includes the
determination of characteristic length from the film thickness and strip
domain period. (4. 5. 6) Bubble domain diameters are routinely measured
at the stability limits but such results are of more limited accuracy.
a^M^...;..^.^^.-..^...,.-:^^^.^^)^;--...—■:^.-. ■ ^v^.^..^:..^^...^--^.^^.....,-.^--..-.^.^--^.-- - ■..■■■.■■. --. .■-^■i-',-,^..... ^-..i^w.:--. .•.■~.--... ...■:-.-.J.t.A1-...j,... .^■■..■. ,.■.,... . - ■ , -:■■. ■■., ..-^.fv. .i. -,,..■ . . .. ... .. ..-.■■. *..,.:../i.::;-\.r.,..*.i^i.**: .:
2.4 Saturation Magnetization and Magnetic Fields
The saturation magnetization was derived from the measurement of
the highest bias field at which normal bubbles are stable (the collapse field)
combined with a knowledge of the zero field strip domain width and the thick-
ness. The theory required has been worked out by Thiele (7), and Fowlis and
CcPeland(5) with useful graphical results in several places. (4) A calibrated
bias coil mounted on the stage of a polarizing microscope served as source
for most of the magnetic fields needed in the work. For situations in which
the sample was inaccessible, as in a variable temperature stage, the satura-
tion magnetization was usually derived from an optical hysteresis loop. (4. 6)
2. 5 Coercivity
A measure of sample coercivity results from determination of the
polarized light intensity modulation resulting from several amplitudes of
bias field modulation. The plot of light modulation vs. field modulation
amplitudes yields an approximately linear relationship which extrapolates,
for zero light modulation, to a finite field amplitude--the coercivity field. (4)
This technique, while a worthwhile relative measure, usually yields values
smaller than the half-width of the complete hysteresis loop and so has little
absolute meaning. Experimentally, samples of coercivity greater than
approximately 0. 5 Oe show measurably degraded characteristics in devices.
2.6 Anisotropy
The measurement of magnetic anisotropy field. HA, by magneto-
optical techniques, was just under development at the time of preparation
of ref. 4 . The recent papers of Shumate, et al. (8' 9) place this measure-
(^a-Jl^ai.■A-^,^^A...^■..-■,.:.^,■^■.v■-.L■.^-^^^^.^^:-.i>^:^t^«Ltl:^^^.f^-..t.,:, .:.,... ^J^-^||i,->,j^tr^,(.^^^^.^|;i|«Va^lf'>1^'"" ... . , ,-.,. ^.^.■...-.:,v.v,.-V..: ..■ .^.-.^ >, : ■-, ■ .. , ■.... .■^.-1.i.».,^.^^;^.'.^,^.i.^..... ■ . .^ < ■ ,:r..:..L^..MHd
ment on a much better footing than it was at that time. In particular they
form the basis for a technique proposed by Josephs'^/ and now used almost
exclusively in this laboratory. In this method a large field is applied in the
plane of the film, rotating the magnetization into this plane. An alternating
bias field normal to the film still yields a small optical modulation propor-
tional to the susceptibility in the normal direction. For fields H greater
than HA - 4 TT MS( this susceptibility is given by X = MS/(H + 4TTMS - H^).
Thus a. plot of X"1 (or the proportional inverse light modulation amplitude)
versus H yields a linear relationship with an intercept at X"1 = 0 of
H = HA - 4 TT MS. This linear relationship is quite consistently observed
when the proper alignment of the sample has been established. The
resulting values of H^ are accurate to roughly ± 10%.
It should be noted that the method outlined in ref. 4 (the assignment
of the low field shoulder as H^ - 4rrMs) is in error. This shoulder is, to
within ~10%, et.iial to H^. but its interpretation is not on as sound theoreti-
cal footing as is the method outlined above, so that the latter has been
adopted as standard.
Z. 7 Temperature Effects
The means of measuring the Neel Temperature using a microscope
hot stage and optical detection has been discussed previously. '4' Compensa-
tion temperatures were measured in somewhat similar fashion where,
however, no microscope was used. Instead, light was conducted to th^
sample by means of a fiber optic "light pipe", then polarized, sent through the
sample, analyzed, and brought out to a photomultiplier via another light
Ihilntniri- - ■ — —mmilWwhMlMiimniiiriififiiiiiiir nniiifiiilMinril ■ ' • "— -' "—"-""—nn i i ir ■>-■-,..*— ■■- ■■ - ■ ■ , , »
pipe. The sample was in a coil within a Dewar vessel where its temperature
was adjusted by means of location in the Dewar and a heater coil. The coil
provided an alternating bias field and the synchronous optical signal served
as a measure of the state of the sample. At compensation the phase of the
optical signal changes by 180°, generally accompanied by a range of high
coercivity and therefore small or zero signal. A substantial drive field was
applied to keep this range <, 10oC and the compensation temperature was
taken as the center of this range.
Determination of temperature effects on characteristic length,
saturation magnetization, and mobility near room temperature involved the
use of a thermoelectric variable temperature stage in the microscope.
Such an addition made these measurements possible over a range from -10
to +60oC.
-'■—■ i iflilHi«0-'^—*" '"-- .w..„-^ ^.„ ■...->■..-.,...■.^ ■...,.--■. -—Mrtiin. ii ir-i
^^^^^^^^^^P^^S«ffiH ■ ■ ~"^
3. 0 LIQUID PHASE EPITAXY
3- 1 Review
Six different film compositions we. e supplied to the Sponsor during
the first phase of this program:
1. Euo.6Er2.4Fe4.4Gao.6Oi2
2. Gdo.8Er2-2Fe4i6Ga04012
3. Gdo.5Y2,5Fe4Ga1012
4. Gd,Yb0 gYj iFe4 2Ga0 8012
5. Gd0-44La0 44Y2 52Fe4GaiO; 2
6. Gdj jTmo gY1Fe4i3Ga0 7Oi2
Although all these compositions were tailored to support stable 5 - 7 mn
bubble domains, they exhibit a wide range and variety of magnetic properties
The anisotropy of compositions 1 and Z is primarily growth induced; that of
composition 3 is stress-induced, while that of compositions 4, 5, and 6 is
probably both growth- and stress-induced. The magnitude of the aniso-
tropies range from about 500 Oe (for composition 3) to 3ö00 Oe (composition
1) while domain mobilities range from less than 100 cm/sec/Oe (composition
1) to over 600 cm/sec/Oe (composition 4).
Because of their range of properties these first compositions were
valuable for basic property studies and preliminary device work. However,
none of the cornpo-iitions possessed the combination of high mobility and
temperature stability that is required for a practical device. Therefore,
during the second phase of the program attention was directed toward
developing compositions with adequate temperature stability which also have
- -- - ■ ■ "-• ^.^^-^^.^^-^ -- -■ -J^.t:,.-....:.....- ...:... ~. ;.,.......-... -
w(j"]!g™ii.w«»www»wMs« 'mmi&mmim&'mGmm?'!'*™?*''™?^^^
the desirable features of several of the compositions in tho above list. The
guidelines for compositional development and the requirements for tempera-
ture stability were discussed in detail in the Semi-Annual Technical Report. (2)
Briefly, the discussion pointed out the potentialities inherent in the mixed
Eu-Y. Sm-Y, and Nd-Y magnetic garnets, and these systems were investi-
gated during the second phase of the program. Since Eu exhibits little or no
angular momentum and Eu3Fe5012 has no compensation point, the Eu-Y
system was of particular interest. Magnetic garnet films of the Eu-Y system
were supplied to the Sponsor early in this program. More recently, Sm-Y
garnets were prepared and supplied to the Sponsor. In addition small
additions of Yb were introduced into the Eu-Y garnets to improve domain
wall mobility characteristics and samples of these materials were also
supplied to the Sponsor. The Nd-Y system was also investigated but, in
this case, the magnetization was found to be in the plane of the film.
The preparation and properties of all these magnetic films are
described in this section of the report. Some of this work was presented
previously in the Semi-Annual Technical Report. U) it is repeated here for
the sake of completeness.
3. 2 Experimental
The apparatus and methods used to grow magnetic garnet films by
LPE have been described previously. U. 2) The magnetic bubble films we]
grown by the isothermal dipping method^ 1) from PbO-B^ solutions. The
films were grown both statically and with rotation. However, during this
;re
,,, ,, , i ,, i.nttai'^""^"'--^- ■■■ — ^- ■■ ■■■-'■- ■ ■ ^ ■-.-
■•T~-™ ^Tr^Ir^,Wf^TO?^fl?rT*-^r^^-W^T-*W'T-7*TF^t'^^
phase of the program most films were grown horizontally at a rotatioii rate
of ZOO rpm to ensure thickness uniformity.
The work was directed toward developing solution compositions in
which could be grown magnetic films which supported stable 5-7 |am bubbles.
In practice, a series of films with constant rare earth ratio but with varying
Ga content was prepared by adding appropriate increments of Ga203 to the
solution. The magnetic properties of the films were measured and the films
were examined for evidence of excessive tension (cracking) or compression
(faceting). Those solution compositions which produced sound, smooth films
having the desired magnetic properties were then used to grow films for
delivery to the Sponsor.
All films were deposited on Syton-polished, (111) GGG substrates.
The thickness of the films ranged from about 3 to lOfam. The thickness
uniformity of films grown with axial rotation was about ± 1 percent.
3.3 The Eu-Y-Yb System
3. 3. 1 Ircroducti'jn
The preparation of E^YjFeg.xAlxO^ and Eu0-6Y2#4Fe5_xGaxO12 films
were described in the Semi-Annual Technical Report. '^J The high Eu, Al-
containing films were characterized by high coercivities («0. 5 Oe) and low
domain wall mobilities (<100 cm/sec/Oe in zero in-plane field). The low Eu,
Ga-containing films also exhibited relatively high coercivities («0. 3 Oe) and
domain wall mobilities of about 200 cm/sec/Oe in zero in-plane field. This
value is about the minimum acceptable mobility for magnetic bubble devices.
Bonner, et al. ^ ' found that small additions of Yb would reduce the dynamic
»KJIBilSWllWM'WMWW' -J ■-,'-' •• ■—'-■' -■-■■■■ ■' ■- ■ '■■■■•■■■■
lUIHUUMIIMIIiilii«! «WH».!!.»..™. .iii>iiMM>iuiiiiilMiu.i|ii|injiiiii jaiiiivniiuiiiiii. ii i ■ iiiaj WII uiuwip^a IIIULIIIIP j i. IUIUU iiiiin)$piim!M«WiniR««^««imiM«P9m^m«linH|
instabilities which give rise to velocity saturation at device magnitude drive
fields and concomitant low mobility values in Eu-Y garnet films. In other
work at Monsanto, ^^' the effect of Yb additions to the low Eu, Ga-containing
garnet films had been investigated in detail. Because of wide variability
from sample to sample, the dependence of mobility on Yb content was not
clear-cut; however, a definite improvement was obtained. The results
suggested that a maximum mobility, in zero bias field, of about 600 cm/sec/Oe
was attained with about 0.2 Yb ion molecule. For in-plane fields of
about 40 Oe, a maximum mobility of about 1500 cm/sec/Oe was attained with
about 0. 1 Yb ion/molecule. Therefore, an addition of about 0. 15 ion molecule
appeared to be a desir'ble Yb content. Having decided the Yb content, the
Eu, Y and Ga contents were then adjusted to yield smooth, unerased films
which supported stable 5 - 7 |j,m bubbles. The dependence of magnetic pro-
perties on compositions of these films as wel] as previously prepared Eu-Y
films are presented here.
3. 3. 2 Properties of the Eu-Yb-Y System
3. 3. 2. 1 Magnetization
The dependence of room temperature saturation magnetization on
gallium content is shown in Fig. 1. The magnetization of other Eu-Y composi-
tions (prepared previously) are also shown for comparison. Magnetization
decrei-ses with increasing Ga (or Al) but does not depend strongly on the
concentration ratio of rare-earths.
10
.■■.... -.^.^ :. 1 fJlllflll-l t— ■■ ^lA.-..^.^*.-~^^>J^^-.^.1.^^... «>.:.. .ft.^.^..-.* i^- ^—•'l rill in I I - — ■- ..-Jw...,^ —..-...- ■ ■ -^.^■^■. ..■J.,.. ^ ...-..■,-■
''•TWW'smiliaiiWJM.u.liJij.wv!«™^^
CSJ «—1
O
CM CSJ CNJ X 1—1 o O O
X
X X X LTS (a (U < O O U.
X 1
X X 1
CVJ U. U. u. >-
LTN
1—4
> CM" - CM >- >-
T o ir\ CVJ
3 O
3 d O
3
/
/ /
< < /<
/ •/
< D D / «a
/
/
/
o o o CD CVJ
rr\
o C3D
CVJ ^
i—i c o
X
o
I d
w o
(Ü
0 u
Ü
m > Ö 0
■ r-l
nJ N
•i-i 4.»
00
0)
00
fc
ssneo 'uojjBZjieußew
u
—•— -- • - —-———— ■. ■- ■-... .■ ■■ -.■.,..■■... -...-.u-:..^:^ ... .,-: .^.^...J.., :„., .y...
rviAiM.w»i.w*4Trwrvr^-^-,W'r^^*^^
3. 3, 2. 2 Characteristic Length
Characteristic length increases with Ga content as is shown in Fig. 2.
It can be seen from Figs. \ and 2 that the magnetization of films which support
5 - 7 ^im bubbles is about 150 gauss and x is approximately 1. 2 ion/molecule.
The average temperature coefficient of the characteristic length from
0 to 50oC for Eu0#5Yb(U5Y2i3:;Fe3>8Ga1#2Oi2 films (the composition delivered
to the Sponsor) is -0.47 percent per degree.
3. 3. 2. 3 Anisotropy
The uniaxial anisotropy field of Eu-Yb-Y films delivered to the
Sponsor is about 900 Oe, which results in a quality factor of about 6 for this
material. The anisotropy field is somewhat smaller than that of Yb-free,
Eu-Y films grown previously (1200-1400 Oe).
Measurements of the Neel temperature reveal an interesting feature
of films grown in tension which is believed to be related to the magnetic
anisotropy of these garnets. In this work, the Neel temperature is measured
by use of an optical hysteresigraph. (4) The signal for a "normal" sample is
illustrated in curve 'a' of Fig. 3. The Neel temperature id marked by the
sharp drop of the signal to zero as the wall susceptibility drops to zero at TN.
For films grown in tension, the curves display an anamolous "tail" or "peak"
as is shown in curves 'b' and 'c', respectively, of Fig. 3. In these cases, the
Neel temperature is taken as the point at which the signal finally reaches zero.
Since this effect is found only in Eu-containing films grown in tension, it is
believed to be related to a component of anisotropy, probably stress-induced,
which is not perpendicular to the plane of the film. Such a stress-induced
12
mm*. v-^>-^^"'^^--^---—■^"■'■-^^^ ...,-.=.i.-,....^.,>,.^.>Jj.,^-^...^..^.-.....^...u^A, .^^J^..^,^_.,Y..J ....^
■..w,,..,,..,,,, ,. ...J.-I.»^... i i < mi upavmwM^'wi'iu-m'« H ~"^m^^Mm^«nWmnaMnit«i i^mmi™*m*m*
1.0
0.8
Eu2 Y1 Fe5-x Alx 012
a Eu
E
B 0.6
to
a> ■^ o «3
o
0.4-
0-4 Y26 Fe5.x Gax 012
EU0.6 Y2.4 Fe5-x Gax 012
EU0.5 Yb0.15 Y2.35 Fe5-X G\ 012
/
0.2
/ /
o
/
0 0.9 1.0
1 1.1 1.2
X , Ion/Molecule
1.3 1.4
Figure 2. Characteristic Length vs. Gallium Content for Eu-Yb-Y Films
13
-■■■-■' 1 ■:■ j^. ■ ■;■■-. ;->,.,■.:...-:.-.■■■-j .L ■■■ .,.,.-■■■ ..J „^ ..■,. -L J":— ■■-•■■■ - - - ■ ■' ^ ■■•■ '■■ ■■■-■-' ■■ ■ warrifl>iM<llitlMI"ilftrlriYili'ii'
.-...„.■.,..„■.:,. ...o^,,.....
"«w.MWw*w^ww^iJi.™i"\!Mimwww^
E >-. 1_ (TJ l_
3 <:
a- o 1_ CD
<
o o
Thermocouple Voltage, mV
Figure 3. Optical Hysteresigraph Tracings of Neel Temperatur« Determinations of Eu-Yb-Y Films
14
JUK.1;.^....;:.^..-...;.-.. .i-:-.-i...^.>^,.:^..^i-».^tl^.J.-.-.^.-..u A<j*c-. .v.j^^i^.^^^^^.^^^-.i.^^itij^^Am^^^^ -■■■'■■" ■-■•-■-■■■■ --■■ ■■ - ■
,— „.„„ivn,,,,,... , „.. ' - .1.1. I.. M»<^*II-*UW««JI»<"<>< "■•^■'■••"'j^w* >!■■ jiiviaiiupimiiiai i.ui q.iwa IUIMITI ix. i» < 11 iiwwnwKl ■Mil II . Jill immiimi Jin w» ■XIIIIIJ lllliup
component could be a result of the large, positive < 100> magnetostriction
coefficient of the Eu ion. If such a component is present it may affect the
operating margins of bubble devices. Therefore, it is advisable to grow
Eu-containing xilms with a close lattice match to the substrate.
3.3.2.4 . Coercivity
The coercivity of the films grown for delivery was about 0.2 - 0.3 Oe.
Such a relatively high value of coercivity appears to be characteristic of Eu-
containing films grown in tension. Lower coercivities can be obtained by
insuring a close lattice match between film and substrate.
3.3.2.5 Mobility
Mobility appears to be quite variable, both within one sample and
from sample to sample of a given growth series. This effect has been
previously reported by Vella-Coleiro'14^ and has been thoroughly verified
in the present study. While no definitive correlation with other properties
has been found, certain trends are detectable.
The response of EuY garnets containing small amounts of Yb to a
step change in bias field is exemplified in Fig. 4. The response at zero in-
plane field is characteristic of an overdamped harmonic oscillator and is
treated is discussed in reference 4. Upon application of an in-plane field,
Hjp of 40 Oe or more, heavily darapled oscillations are observed which per-
mit analysis of the type discuosed in reference 15. The frequency of these
oscillations generally increases as an in-plane field is applied while the
dec^y time exhibits a tendency to decrease, as is illustrated in Fig. 5.
15
'^(JIJ»ww^»^w-i.^*^'.-,'fj^^A^^^ ■ -^a^rswffltrvTFT^wim^riw^^wpi'^wwwrw^^
na d
.5? in
o. O
20 40
Time (nsec)
60 80
TO c
CO
(TJ o Q. o
Time (nsec)
Figure 4. Optical Response of a Typical Sample of Eu-Yb-Y Garnet to a Step Rise in Bias Field at Various In-Plane Fields
16
-- -■ i
mmmmmmmmmmmm^m-mmmifrmmmmmmmmmmmmm wmmmmmm. mmmmmmmmmmmm:s
40
30
Oi
3 20 ■o
10
0
40
• •^,
i I I I i I I I I I I
s
S
30 ISI
20
/
s
• y s
s s s
s s s s s s s*
10 l_L J I I L J I L 50 100
Hx (Oersteds)
Figure 5. Natural Frequency and Decay Time vs. In-Plane Field for a Typical Sample of Eu-Yb-Y Garnet
17
A.'fcL-iluL/.^Lfe^JuL . ...^ ^..-^-.^^ .^ -■ ^- - - — ^— ,-^..^_ ..-.^ A, . _. ti',illK-n*rfWall.-......:! .■ ..^.ti^;' ^-r.ru!
wmm*mmmmm*mmf^<mmu.mimi™mmmmmmm*miimimm^*m***^*^mim*mm^*mmmm*t mmmmmmmmmmmmmmmmmmmmm
is con- There is a reasonably linear frequency vs. Hx characteristic which u
sistently observed in all compositions in which this oscillatory behavior is
observed. The internal agreement of the Td data must be considered largely
fortuitous.
Fig. 6 illustrates the behavior of the mobility with in-plane field as
determined both from the oscillatory behavior and the over-damped results.
where these have been observed. The oscillatory result at Hx = 0 is
determined for the extrapolations shown on Fig. 5. We tentatively assign
the difference between these results to a suppression of turbulent motion of
the wall by the in-plane field. W The presence of an in-plane field in a
device will clearly improve the speed of the device.
The variation of mobility with temperature has not been a routine
measurement in this study because of the various equipment complexities
involved. It is, nonetheless, an important characteristic of any material
destined for use over a significant temperature range. We have therefore
recently initiated such measurements using the step field response method
within a thermoelectric variable temperature stage.
Mobility was found to decrease as temperature was lowered, as is
shown in Fig. 7. The wall mobility in Hx = 0 at -100C ranges from 50 to
150 cm/sec/Oe. For Hx = 33 Oe (limited by the variable temperature
apparatus) low temperature mobilities between 180 and 500 cm/sec Oe were
observed, there being a definite correlation between values for H =0 and
those forHx = 33 Oe.
18
.. , . ■.■....■.■.-J..^^i„»^J^^-.-^.-,.J—. „^.x^^x^.— L^^^^-a.^-.»^.!^-^^.!^^.^^, . _,. . ___^, -*
"■""H.I.I.".!! •"•".. T—^- -.1 .1 >«uKi>aai»iii •jijiii.iii..,jiii>iipilwiii i ii ü. i na« iiiuau,!. .I ji j ii ui i uw^mmBii^Minqmmnm^mngi
2000 / /
a> 1500 O
• y
CO y
b y*
^ 1000 y
y
y 500
\
(
n i i i i 1
y
J i i i J L 50 100
H (Oe) x
• - Results from Oscillatory Curves and (at H = 0) an Extrapolation of those Results
0 - Results From Cverdamjjed Response Curves
Figure 6, Domain Wall Mobility vs. In-Plane Field for a Typical Sample of Eu-Yb-Y Garnet
19
U^ygjUgy,, .J„J..J..„^^:.....„:..^.:^-^^ .^.-^-^^^ .... M| - ,, , ■ " .■.■^..,-
>7)PVHHR^vR3(3p^t<rri>np^
600 -
500
H - 33 Oe A
400
03 O
a» | 300
/
/ / / /
200
/ /
/ /
100
0
__11J_1 __ ^•—•
-10 10 20 30 40 50 60
T(0C)
Figure 7. Domain Wall Mobility vs. Temperature in a Typical Sample of Eu-Yb-Y Garnet - All Results are Over- damped Response Curves
20
i^- - - - —- -■ - ■- ' ' ' -"■-^-— - -^-
'v*^mi*wr*mwJ!fv*r*v.mm•*'wim»^v^*y9n*v*m'jrm' jmull»*™mmKm^ymmm■jflWiJ.'i«"WAWJM,i"iu'iiiwiM^nuniifMmnj(iulup,«!)^ »iwi^pl"^«*iH»^jiPHPW**,"^mw^^WEW*W«^«J'll!Wip^W"S^l*|.|l4W.w»Wlmimmmm**m&'VmmiVimmm}i*mii
There are some interesting correlations between low coercivity and
high mobility values within this system which should be further studied.
Possibly both of these desirable properties are associated with the absence
of a very small scale network of defects. Such a network, when present,
might not only hold up domain walls, thereby increasing coercivity. but
also cause the wall motion to be more turbulent and thus reduce mobility,
perhaps through nucleation of in-plane Bloch lines. (17' 18) This netWork
might arise at the film-subs träte interface from a lattice mismatch of
substrate and film too small to produce cracking.
3.4 The Sm-Y System
3.4.1 Introduction
It has been pointed out previously^2) that Sm ions are among those
which are of interest as additions to Y3Fe5_xGaxOi2. Sin3Fe50lz, like
EujFegO^, does not exhibit a compensation point. Hence, the magnetic pro-
perties of the mixed Sm-Y garnets are expected to exhibit good temperature
stability. However, the damping parameter of the Sm ion is much higher than
that of the Eu ion^ ' and, therefore, might be expected to affect the intrinsic
mobility of Y3Fe5_xGaxOj2 films to a greater extent than Eu ions. On the
other hand, since some damping is required to reduce dynamic instabilities
and to linearize the velocity-drive field relationship, mixed Sm-Y garnets
(containing small amounts of Sm) might be expected to exhibit reasonable wall
mobilities at practical drive fields. Therefore SmyY3_yFe5.xGax012 films
were among those grown and supplied to the Sponsor.
21
J........... ■,.,.- ;.......^,.-^.-.J...,.:...;......J.^_..■■...... „..,..,.„-,...:. ^__^ ^^.^
~ J "• W^*-^»^"^'*Wi^,'*-'!"'!^r-'iT»,-"J-S^IMJv.«■H^ririH':iw^!f'/vi)-t»ij,fl^fwij^.lmM.>W.,iiNi",«!"W^i.rxM.*i.LM.iJ.xhwr.rai»4viH^J^^,l^^^.(T*^!.n^i^^.^L^UM, i.i|»f!i«w'ir.':"1.'-'!M'iWIWJ'-'PTWfT.' 1*n^„r»»iFJui,(^»i^?.■*»,-«(u.*f ™.*m «w■ iu-tn I»JWW,*I»IM'JHifl-ii.-» »Ä^nwpw«^-7«
/
/
o<o /
/ <
o
/
a/ /
/
/
/ /DO
vO ^«a- CSJ oo l-H CSJ crv CO
o •
o O o E E E E oo CO CO in
<j D 0 •
±
ro C5 C3 CSJ
C3 O
cr\
a» csj u _: ^
E 13
"to O
o
C
Ü
in u o
ö a)
o U
S 3
(t) Ü
> Ö 0
.f-4 ■u
« -w (U ä
00
ssneo 'uouezjiaußBW
22
- i — — m~M <mm -■- ■ -
•-^■•TO»p»r^piWFijw«HrTT"w^^ "•Wir wff," •.WVV»KIUJ,JJ*'IPH <iwww^*w*^iww*jwr»r*mmvw^vwwv'enviv*w
3.4.2 Properties of Sm-Y Garnet Films
3.4.2.1 Saturation Magnetization
In Fig. 8, room temperature saturation magnetization is plotted as a
function of gallium content for films containing various concentration ratios
of Sm/Y. Since the films are not of the same or optimum thickness, there
is some scatter in the data. However, the genera}, trend is obvious.
Magnetization decreases with increasing Ga and does not depend strongly,
if at all, on the concentration ratio of rare-earth ions,
3.4.2.2 Claracteristic Length
The behavior of the characteristic length, I , with composition is
illustrated in Fig. 9. Again, the scatter in the data results from the films
being of different thicknesses. Characteristic length is seen to be indepen-
dent of Sm/Y concentration ratio.
Since the characteristic length of domains in films which support
stable 6|am bubbles is about 0. 75 jam, a Ga content of about 1. 3 ion/molecule
is required.
3. 4. 2. 3 Coercivity
Fig. 10 is a plot of coercivity vs Ga content of the various films.
The coercivity is quite low and independent of Ga content for x less than about
1.2 ion/molecule. For higher x, it increases rapidly with Ga content. This
behavior is believed to be related to the lattice mismatch between film and
substrate. The films are grown in tension and lattice mismatch increases
with increasing Ga. The rise in coercivity with Ga is believed to result from
23
.. : ■ ■-■■ ■ — ^ ■ ■■».^■..,M...^.....1.<..--„,...J^...^I.....,..,.,,...., ■ , ■..,a......«:,^^.^ ~"~"Tr"'",^l''*^J'^^°JWa**'>*>^
*■— — ■-i-w w ■nFiK in ..«■■■.i. .w u.iiiii ■•■ I.M. RiBMiimm ■■■ i i .iiiiii.iijii«iH^«nKmnv^g^a> ■ "l ' i ■■11 ■>! mMMiim i 'mm IIJIIII.
V \° D
> ■HI
\a
•\c < \ c )
\ - i > OD
\
\ D
\ -
o
SO "^ CVJ NO ■—« CSJ m rr\
o o o o E E E E
CO LO l/l CO
< a o •
1 1 >-L 1
m
CVJ
o oo NO
o
0) Ö M cd Ü
fH | s
00 IH 0
CD m 4->
zs Ö o (U 03 ß o 0
^ U
c a o
«■MB i—i ■-I
•« rt Ü
a) • -t"» tn c > o O
&0 E Ö =3
mmam
lü u .rJ
O to • »-i
u (1) -u U (D U fit
rC u
• ON
to u 3 00
h CM
C3
Hill 'LUßUBI D!;SU9PBJBM3
24
ki.!!;^.»^.^..;.: .,..l.,.,>.,..^-..i^^v...-->.v..-.^l. ■Ji,^.^^J.--.^.-.1:.J-.--A^'i:>^ciJl... ^^^,^^^:'^..:^.^:l^..lJIJ^-i^;..i^^lkv^ .■.w.^J-llf«^^.^.^.^^.^.,, ..,.,. . ^ ,.., _.,,, ,, ,., .„. , ji ■■.■■..-■.....s ^L.;.,.:-... i-jtm i».-:.:■I-.»-:.*;».;
-^wnuJilW^Hi'W^W'mMigvwJulim^^^
13 O CD
o
o o
0) Ö u
Ü
I
CO
u o
t! o U
a
Ü
(D >
u h 0) o U
0)
SO
h
90 'AUAIDJ900
L 25
- iinMini Itrifiiiiiiiiiiiritoiff^WiMiaif^ i ■ ■ -.. . ..:....:..-..._w^.V,.i.;...
■■--J til*«,:,. , ......J/, ^**^«»^irt»4UÄw«ii^.ii^ai^
ni>™^.™»"<l'»"W'Il,HMMUPMW!W^^
excessive mismatch. Since a Ga content of about 1. 3 ion/molecule is required
for stable 6 Urn bubbles a good lattice match could be ensured with a Sm
content of about 0. 5 ion/molecule (based on Vegard's law).
3.4.2.4 Anisotropy
The anisotropy field and the anisotropy energy density for a number
of films are given in Table 1. These films were grown at somewhat different
temperatures and growth rates and since anisotropy is dependent on these
growth conditions, the values are not strictly comparable. However, the
values are illustrative of the magnitude of the anisotropy which can be
obtained in this garnet system. The anisotropy field of films which support
stable 6 ym bubbles is about 1500 Oe and the saturation magnetization is
about 150 Gauss. These values result in a quality factor of about 10 for
these films.
TABLE 1
Uniaxial Anisotropy of Some Sm-Y Films
Anisotropy Anisotropy
Sample No. Comp Sm
0.16 0.16
Dsition Ga
1.23 1.28
Magnetization (Gauss)
220 146
Field (Oe)
Energy Densi (erg/cmz)
5564A 5565A
970 1480
8468 17150
5567A 5571B
0.24 0.24
1.25 1.31
183 134
1220 1440
8859 7657
5577B 5578A 5586C
0.32 0.32 0.32
1.18 1.24 1.27
228 194 149
1210 1540 1500
10948 11855 8870
5562B 5562E
0.36 0.36
1.25 1.33
187 134
1770 1685
13134 8960
26
M*****^-^^**^^. . - ,. ^ „, ^ ■^■._ ^..^...,_J. _. ..
■-»■W-STP-W'^W^^^ 1zwzrFmxrmfaqmrrvw&^^ *m
The data of Table 1 suggest that the anisotropy energy density
decreases slightly with increasing Ga for a particular Sm/Y concentration
ratio even though the anisotropy field appears to increase. This is probably
a consequence of the rapid decrease of magnetization with Ga content (Fig. 8).
3.4.2.5 Mobility
Sm-Y films, when subjected to a step increase of bias field, always
exhibit a simple wall motion to the final position (usually exponential) rather
than an oscillatory motion. This is in keeping with the higher damping to be
expected from Sm ions compared with Eu or Yb. There are additional
problems present, however, which are presumably associated with the pre-
sence of Bloch lines in the domain walls. A typical sample which has been
subjected to a variety of previous treatments will exhibit a characteristic
response time (to 63% of final signal) of ~ 200 nsec. at Hx = 0. During this
measurement the stripe domain pattern is observed to slowly move and
change form. A similar slow response is often observed at Hx = 40 Oe.
However, when the sample is subjected to a large in-plane field or a
saturating bias field and then returned to Hx = 40 (not going through zero) a
response approximately an order of magnitude faster is observed. In this
condition the stripe pattern does not move visually. A reasonable explanation
is that Bloch lines, originally in the walls, have been disposed of and the
latter response is charac'.eristic of walls free of Bloch lines. If Hx is now
returned to zero one can often observe the pattern begin to move, first in an
edge of the field of view and gradually spreading throughout as Bloch lines
27
-■ ■■ ..„ J. u ^.^.„„.„^.^„„.^^.^^ ^ .^ -■■• ^^Ä«^«fii«»^siiua|j||^^
^^n^*?*^WfWns?WPp^T?rTWW*r^^ IWTI ^U^^W^'»"^«^»^"^.^ •?^^»fl^flm5BW?^5W^W7f7?n^W?^^:W^^
spread through the pattern. One would expect that during this process a
combination of the fast and slow responses might be seen. This has indeed
been observed. In what follows emphasis is given to the faster response as
that to be expected in a film in which hard bubbles have been suppressed
(for example, by ion implantation).
The mobility values versus in-plane field are given for three compo-
sitions in Fig. 11. From the foregoing discussion it is clear that the
values near Hx = 0 may be depressed by the presence of Bloch lines. How-
ever by Hx = 40 Oe this complication should be absent and, indeed, the data
show the qualitative reduction of mobility to be expected as more of the
highly damping Sm ion is introduced. Variability within a composition
remains a problem with other samples of the x = 0. 32 composition showing
mobility at Hx = 40 as low as 180 cm/sec/Oe. It is clear, however, that
quite useful mobilities are possible in this composition.
3.4.2.6 Temperature Coefficient of Characteristic Length
The characteristic length was measured as a function of temperature
from about 0 to 50oC and the average temperature coefficient of characteris-
tic length (100^ £/£ AT) was calculated for that temperature range. The
coefficient was about 0. 5 percent/degree for all the compositions studied.
3.5 Nd-Y System
Epitaxial films of NdyY3_yFe5_xGax012 were deposited on (ill)
Gd3Ga5012 substrates. The nominal value of Nd ranged from 0.4 to 0.67 and
x ranged from 1. 0 to 1.2. The films were smooth and were not cracked.
28
am,;-:.:.^.^^- .... -^.. - ■.,^.J-.,..„.;-.,J^-^...^,.....-. „.J.J,^a..^..^^^>...,.....„.,J^„^^........,^.^,....„ni„.^..v..^.-. ■ .,..,.-.„ ...„. >.. ^.. „ _ . _.„_ ■^...„:^....,.,..„ ..,
. in HU ■iimi.ji in ..,^w-.....«i.i.u«~—T, j iLi,iiiuia..iini .H i niii > jaannn.Baai u! i. n IIKüI MPIUMI 11 11 u . » ji. i •. > 11 jwiii uanpHnnnininmn«^^^«'
10C0
o
o CO
E o
J3 O
800
600
5 400 -
200
/
/
/ /
/ A
/ /
/
/ / / /Y = 0.16
X =1.2
/
Y = 0.32 X = 1.3
/ /
/
4 • Y = 0.36
Xrl.3
I 50 100
In - Plane Field, Oe
150 200
Figure 11. Mobility vs. In-Plane Field for Some Sm,rYq „Fe, GaxO, „ Garnets
1yI3-yr e5-x
29
.^^a^,^...^^:-.-.■■...— ^^^...■^■^^.■i.^ ^^.i^J...^<V^*^*i)llt]iföa±i^ ■-■lVt-^..^i.^W^.^,J^^^..,.^,..;^,.J... .... ;„., .■..,-„■..■......- „;■..■ ..-..■.:■■:.,:... A.^-^^^.^:,.^.^...
Thicknesses ranged from about 3 to 20|-im. The films did not exhibit uniaxial
anisotropy in the < 111 > direction, the magnetization apparently lying in the
plane in all c?,ses. Anamolous anisotropy effects have also been noted in
bulk, flux-grown, Nd-containing garnets by Gyorgy, etal.v1^'
3.6 Discussion
Both the Eu-Yb-Y and Sm-Y systems appear to be promising candi-
dates for bubble device application. The garnets exhibit reasonable
mobilities, satisfactory quality factors and good temperature stability.
Coercivity can be minimized by assuring a close lattice match between film
and substrate. Compositions which support stable 5 - 7 Mm bubbles have
been identified.
Additional work with these systems could be quite rewarding. The
relationship between mobility and composition is not entirely clear. In
addition the proposed relationship between mobility and coercivity particu-
larly apparent in the Eu-Yb-Y system deserves further study. Such a rela-
tionship could have an important bearing on bubble technology regardless of
film composition. A detailed study might lead to a better understanding of
wall mobility and wall turbulence in garnet films.
30
^atemm uut^^mummmtmmmmmtmmi lüHHlMMMHMMIk ■ ■
4.0 PLASMA SPRAYING
4. 1 Introduction
Plasma spraying was investigated to determine its potential to pro-
duce magnetic garnet films of the composition GdEr2Fe4.5Ga0.5O12 with
superior or unique properties, or to produce magnetic films with satisfactory
characteristics at lower cost than traditional crystal growth processes.
Two approaches were designated for study: (1) the conversion of
plasma-sprayed polycrystalline garnet films into monocrystalline films,
and (2) the direct fabrication of plasma-sprayed monocrystalline garnet films
The first approach was a two-step process for forming monocrystalline
films. It consisted of spraying polycrystalline films containing garnet and
flux and subsequently converting the deposits into monocrystalline garnet
films by a thermal treatment. Results of this work were reported in the
Semi-Annual Technical Report. (2)
The second approach was a simplified process in which the film
deposition and thermal treatment were performed in close order to reduce
the operation to an essentially single step. This method was studied during
the second portion of this contract and results of this work are reported
in this section of the report.
4. 1. 1 Growth from the Melt
The first and most straight-forward approach at garnet film growth
by direct deposition was growth from a melt. Garnet powder was fed
slowly into the D. C. arc plasma torch and deposited as a molten film. This
approach was unsuccessful due, primarily, to the temperature variations
31 Kfai.-..^^.^..,..,...,...
■ ■ ■ ■■ ■ ■ ■■■
experienced by the substrate during plasma spraying and the inability to
maintain the stoichiometric compoeition of the incongruently melting garnet.
Films deposited in this manner were essentially polycrystalline.
4,1.Z Growth from Solution
Other studies were made using a modified liquid epitaxy approach
similar to the work performed during the first period of this contract, except that
both plasma spray deposition and subsequent single crystal film growth
were accomplished in the same apparatus without interim cooling to room
temperature. The heated particulate was sprayed onto the substrate and
formed a molten solution of flux and garnet on the subst^te surface. Film
growth was initiated by cooling the solution to a preselected growth tempera-
ture.
After several conversion runs were made using the D. C plasma
torch as both the means of powder deposition and the substrate heat source,
an auxiliary resistance furnace was found to be a more stable and uniform
high temperature heat source for film growth. Thereafter the D. C. plasma
torch was used only as a means of providing heat to the powder for melting
and deposition on the substrate.
The D. C. plasma heat source was abandoned in favor of the R, F.
plasma source early in the program because of the latter's lower gas
velocity and resulting gentler deposition of the molten particles. In
addition, the absence of consumable electrodes in the R.F. plasma resulted
in a cleaner film growth environment. As with the D. C plasma, the use of
an auxiliary heater during film growth provided a stable and uniform growth
temperature.
32
^k^,^^M.r>1^1rfh^|.^--^^.^^^^^^--^.^--^^^^--..U-^^.,r...| ,nr,,-(|rf;,i„..-^^-^------.-.J-.-^---.---^ •■.■--■-.. :^... -.,.■■ ■■■^-J-,:^i.---... .. ,.■,■■......_.,■ . :....,......,,....... ... ..^..-:........■...■,„.■ :.-.. ■..,... >..M....^Wt/.^M.t.^..O..^^^iü>,Y..>-.:^t.^.^r.^^ .^-^r , ■ niiMÜtf
The best bubble garnet films prepared during the second period of this program
were grown using the R.F. plasma torch in conjunction with an auxiliary heater.
4. 2 Experimental
4.2.1 R.F. Plasma Spraying
The radio-frequency (R.F. ) plasma process used during most of this
study was first proposed by T. B. Reed(20). It represents an attractive way
to heat a particulate to a high temper iture without introducing electrode
impurities. The overall process and equipment are shown schematically
in Fig. 12.
Electrical power is supplied to a load coil by means of a 30-kW
induction heating unit (Lepel, Model T-20-3-DF 1-E-H) having a resonance
frequency of ~ 3. 5 magacycles. The plasma acts as a 1-turn secondary coil
to the 4-turn load coil. A shielded coaxial copper tube is used on the positive
side of the R.F. coil to prevent energy loss.
The R.F. plasma is initiated by a high-voltage discharge produced
with a Tesla coil. Once the argon gas is partially ionized, it couples
efficiently with the R.F. field and sustains itself continuously as long as
plasma-forming gas is supplied. In order to efficiently initiate a plasma, the
impedance of the power unit and of the load coil must be properly matched.
By so doing, the reflected power is substantially reduced.
The R.F. plasma torch (shown in detail in Fig. 13) consists of two con-
centric quartz tubes with three independent gas streams. Cooling air flows
between the tubes to prevent melting. It is blocked from entering the hot
zone by a graphite ring. Argon gas flows through the inner tube and forms
33
am ,.—■..■.■■J-^,..,-.^...,.».—Jj..-.....i,...J.i..;..—..^..,., ^^-^lyiiiiiimiiiiiitniliiniiiih'ihiriiln'iiilt ■■—'— -• m..! r n.m ■ ■■ ■ - ■ - utu . ^ . t_> u
<D ^J '*—» ID CD H- C •a CD CD 00
Li_ 03 i_ O CD cn -♦—' 'O c a> 5 c o
Q_ E c "S £ < 05 8
o CO E C
o 1— o OL CD Q. o
C ) -o ro 1 CD CD ^ 1
a> o
c CD
O o XD TO O
1 O o o
I
CD
CD ■a
on o CD -Q X 13 4±J F
1— ro 3
fsl c u. n •♦-' ro 3 ro 3 O CO Q_
cn t_ OJ c -^ i-
• r; ro I— ro zi — CD -C ro X a> E gQ.|
CD
8 03
— a.
8 o E
ö o
Ö
u
CO
01
PS —^ VH CJI 0 c u ro -« OS X g c 1) o Jl
-I-' u o w 3 u c (M ""• ^-l
o c 03
1)
3 =3 >H
§■ h
34
-- ■■ ■ ■■ "- ■HM» - - - - - -—- —...._. . ■ .. ^
Feed Powder Inlet
J
'O'-Ring
Inner Quartz Tube
[I j Water-Cooled Powder Feed Tube
Rubber Ring
Plasma-Forming Gas Inlet
Outer Quartz Tube
Plasma Flame
Substrate
Platinum Holder
Heating Furnace
Alumina Pedestal
S.S. Stand
O'-Ring
Cooling Air Inlet
R.F. Coils
Graphite Ring
Alumina Tube
nsulation
Furnace Windings
Specimen Thermocouple
S.S. Tube
Thermocouple Wire to Recorder
Figure 13. R. F. Plasma Torch Detail
35
r^»..^.JA,,^-.-...mV.,^..^^^--..^u-^.^..^.,^^.^...^J.a,^M»».^^.^>.JJi^^.^~M-^. IM :-:-^--.^. ■■ .... , . ,. ,.^ •^.«.«„ö^m
the main body of the plasma. It is injected tangentially so as to create a
vortex which forms a low pressure region in the center of the plasma.
Circulation of some of the ionized gas up the center of the plasma provides
enough seed gas to continuously maintain the plasma. Finally, oxygen gasi
flowing through a water-cooled copper tube, carries powder through the high-
temperature plasma for melting and deposition on the pre-heated substrate.
A cylindrical 4-turn R.F. coil of copper tubing surrounds the outer
tube and serves to couple energy into the plasma. To provide a uniform
preheat and growth temperature over the substrate, an auxiliary furnace is
fitted around the outer quartz tube below the induction coil.
After preheating the GGG substrate to a specific temperature in the
auxiliary furnace, the substrate is quickly removed from the furnace and
positioned at the appropriate distance from the powder tube outlet. Powder
is transported from a feeder using an oxygen carrier gas and passed through
the high-temperature region for heating in the plasma stream. Powder is
thus melted in the plasma flame and deposited on the substrate. The sub-
strate temperature is monitored continuously using a Pt, Pt - 13% Rh
reference-grade thermocouple which is in physical contact with the bottom
of the platinum substrate holder. After deposition, the sprayed substrate is
quickly returned to the furnace for film growth and held for a specific time.
Once film growth is terminated, the flux is decanted from the surface and
residual flux is then removed using a dilute solution of HNO3.
36
I ■ -—-■ ^-^~^w^^-—-. ^^—^....U^^ , . ...._. „„., ,,... ., n,.^
4. 2. 2 D. C. Plasma Spraying Station
Initial experiments to fabricate bubble garnet films by spraying
without cooling to room temperature and subsequently reheating to form the
epitaxial film were made using the D C. arc plasma torch discussed in the
Semi-Annual report.^)
The D.C. plasma spraying station is shown in Fig. 14. The plasma
torch is mounted against a metal bellows which allows movement of the spray
torch to permit uniform coverage while minimizing contamination from cir-
culating air during the powder deposition. A quartz tube is placed against
the bottom of the bellows to act as an enclosure for the substrate during
spray deposition and film growth. The auxiliary furnace is placed around
the quartz tube to provide a uniform growth temperature across the substrate.
Once the substrate has reached the preheat temperature, the plasma torch is
activated and powder is injected through the plasma. The molten particulate
then collects on the substrate and the specimen is held at the growth tempera-
ture for a specific time as required to form an epitaxial garnet film. After
growth, excess flux present is removed in a dilute solution of HNO3.
4. 2. 3 Material
Gadolinium oxide (Gd203), erbium oxide (E^C^), and boron oxide
(B2O3) were obtained from Research Inorganic/Organic Corporation and were
99.9%, 99.9%, and 99.99% pure, respectively. Grade I iron oxide (Fe^),
gallium oxide (Ga203) and lead oxide (PbO) manufactured by Johnson
Matthay Chemicals Limited, were obtained from United Mineral and
Chemical Corporation,each having a purity of 99.99%.
37
■ ■ ■
---— - ■ 1 '•--—■—^—^■^^^^■;...„,.-^..„^.^^.^.M^I^^I—^-^..■.^..^^^,^—L^^.^.^^^U.^^. ,.,. ..,...-.■....— _ „__,.... ... _ ....... ■—- --^- 11 rii
Plasma Flame
Substrate
Platinum Holder
--Metal Bellows
Heating Furnace-
Alumina Pedestal
—Plasma-Forming Gas Inlet
— Feed Powder Inlet
Quartz Tube
Alumina Tube
Insulation
Furnace Windings
Specimen Thermocouple
— Thermocouple Wi re to Recorder
Figure 14. D. C. Plasma Spraying Station
38
-■■■'■■.■■■ ■■.■..'■■.. ■...-... .....,-,.■ ....,.J. :-^;-^^- ^^^-^^.^ --.■.^.....-..■,- .; J..^,^u..^-.^^.:^..^:^-..^Y^IlMWlVü""^"^*-^^^■"^■■■-""-^ ^■.-■^■^-■■-J-^W-.-M^.--M. ■..>■ ■ . ..- .....- ...... . ■....■..■.■■■.. .. ■■.■ . , : ,..■■ ;.. ; .v.-... ^ ■ ......1 .. , A ,-....■■■. . ■ ■ ■ - ■ ■■■■■■-
The individual oxides were weighed to four significant figures using
an Ainsworth triple beam balance. The oxides were combined to form the
mixture shown in Table 2 and then blended for one hour.
The blended oxides were placed in a platinum beat and heated for
Z hours at 1300oC in air. The sintered charge was then pulverized into a
fine powder which after sifting, was suitable for injection into the plasma
torch and used in studies to determine the feasibility of producing epitaxial
garnet films directly from the melt.
A homogeneous mixture of the oxides in the garnet and flux powder was
obtained by forming a solution at high temperature. The solution was heated
in a closed platinum crucible at ~ 1100oC for 2 hours and then quickly poured
into a platinum boat. The resultant ingot was pulverized using a mill
containing a tungsten carbide rotating blade to form a powder which, after
sifting, was suitable for injection into the plasma spray torch for producing
films from solutions..
The < 111 > GGG substrates (~ 14 mils thick) were all obtained with
one side polished. Some substrates ^3/4" diameter) were scribed on the
unpolished side and broken into quarters in order to conserve substrate
material. Other (~ 1/2" diameter) substrates were used as received. The
substrate cleaning procedure prior to spraying has been described
previously. '^/
4. 3 Results and Discussion
As in the first portion of this contract, the approximate composition
GdEr2Fe4#5Ga0>5O12 was studied. This garnet exhibits growth-induced
39
TABUE 2
Powder Compositions
Garn< it Pc for
)wder
Growth f rom the Melt Component Weight %
Gd203 4.49
FejjOj 77.66
Er203 13.05
Ga203 4.80
Garnet & Flux Powder for
Growth f rom Solution Component Weight%
PbO 91.26
B203 1.83
Gd203 0.31
Fe203 5.37
Er203 0.90
Ga203 0.33
40
...J..„..^i.J-„.. ....^-^ ...J-^.,.. .,..„■. ^ ■-.:■.. -...,--. ..,...,.,„-■■.....„..^^^ .,....,....^..-..,-...,.....,^..-. ^,^^..^^^. ., „.^....^^^„.V...^...,,.....:......, ..,. .^.:1.....^ ^^.^..,...... ,> ^«a
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^mf^^^^^^^^^mrn^^^^^^tmmmm^mmmm^mr^msmm
anisotropy perpendicular to the film plane. A good lattice constant match
is obtained between this garnet and < 111 > GGG substrate.
4. 3. 1 D. C. Plasma Spraying
4. 3. 1. 1 Epitaxial Crystal Growth from the Melt
Efforts to form single crystal garnet films directly from the melt
were largely unsuccessful. Powder sprayed using the D.C. plasma torch
was slowly deposited on GGG substrates, maintained at 1200 - ISOCC, and
then slowly cooled to room temperature. However, the resultant deposits
were essentially polycrystalline. This was attributed to temperature non-
uniformity and temperature fluctuation within the substrate during spraying.
4. 3. 1. 2 Epitaxial Crystal Growth from Solution
Magnetic gr rnet film growth was achieved by using the D. C. plasma
torch for depositing a film of garnet and flux on a preheated substrate and
subsequently maintaining the specimen at the film growth temperature for
several minutes. However, domains were observed only in isolated areas
of the film after conversion and the results were not reproducible. Typical
conditions used for producing a garnet film by D. C. arc plasma spraying
are shown in Table 3.
In general, the films produced by D. C. arc plasma spraying were
inferior to those produced by the R.F. plasma spraying method. The high
kinetic energy imparted to the injected particulate caused the molten flux
and garnet to flow off the substrate during deposition. Consequently, only
a portion of the sprayed solution was available for growing an epitaxial
film on the GGG substrate.
41
- -~*^^~.^—^. iaMim. . i i— --"■ -- : " *•■ <Si
•*~™rTTr!?*r*>*f^.**w*rv?*^^
TABLE 3
Typical Conditions for Growing Epitaxial Garnet Films from a Flux
Using D. C. Arc Plasma Spraying
Spray Torch
Arc Gas
Arc Current
Plasmadyne Model SG-3 (25 kW)
Argon at 50 cfh
250 amps
Powder Feed Mechanism
Type
Carrier Gas
Plasmadyne Rotofeeder
Oxygen at 15 cfh
Growth Conditions
Substrate Preheat Temp.
Growth Temperature
Growth Time
930oC
930oC
10 minutes
42
i mMüMÜI I ■"•— -''■-'• .^^■■^^»^^^^^.^.^....^^.^..^.s^^^^,^ ,.„......>.«:j^.c.ii,..,iJ,».j.^,-.... ...,..^,.,..„.^.,......„..,,^^d»^a.^c~^^^~^,.»i^^...^-^.^^iBlMl
i ^'■HI»IMWI»H,J*M.IW,.WI^«"J-'^.I'!*J^,1^!«'*..-,1- «w*wiIWir*fJM4W^.r.tM WIJ^ I WiP^W^PtP^^J.WW W»W.^"»W« ^WW-V f^PWiffVSRW ^^''^P',"^W!^w'!swTw,^sTs*rwr^fT^^TrWw^Jr^^T^^
i. 3. 2 R.F. Plasma Spraying
The R.F. plasma method of growing bulk single crystal oxides was
succesifully demonstrated by Reed. (21) However, the oxides in his work
were simple congruent melting compounds in which bulk crystals were
grown directly from the melt. The difficulty in maintaining the correct
composition of the bubble garnet from the melt made growth from a solution
more attractive. However, garnet films made from a solution were
inferior in overall quality as compared to films made by liquid epitaxy using
the "dip process. "
The R.. F. plasma method has several drawbacks when compared to
the liquid phase epitaxy method. First, the temperature uniformity across
the substrate obtainable by the "dip process" is difficult to achieve using
an R.F. plasma. Temperatures vary from the center of the plasma to the
boundary causing composition and film thickness variations in films grown
by R. F plasma.
High defect density can probably also be attributed to the plasma
process. This conclusion is based on a liquid epitaxy "dip" experiment
using garnet-flux prepared by melting the powder employed in R.F. plasma
growth studies. In addition, GGG substrates were prepared in the manner
typically used for spraying studies. The time-temperature conditions pre-
viously found satisfactory for film growth by the "dip" process were used.
Nearly defect-free films were grown indicating that substrate and powder
preparation procedures were satisfactory. Apparently, the exposed surface
of the GGG substrate during and before plasma spraying was vulnerable to
43
.i.i.>v„.......... ,....■-. ; ,.., ,.;,......-.,,-^jc....„-.^—,.^,...,„ ..-,J.J..„.„...^„„-,.....^.Je.,^-..,...:.^....^.,..... ^ ......t,.. ...^^J.,
airborne contaminants. Also, the long exposure time of the substrate to the
flux probably contributed to the high defect density.
Finally, it was difficult to maintain the sprayed solutions in the
garnet stability range. The films were invariably contaminated with
crystallites of orthoferrite, hematite or other phases.
4. 3. 3 Characteristics of Selected Garnet Films
All films produced in this program were qualitatively examined using
an optical microscope to determine defect density and overall film quality.
Domain patterns were observed using polarized light. Film color was
macroscopically noted.
None of the garnet films made by D. C. plasma spraying from the
melt exhibited magnetic domains. However, films primarily sprayed from
solution did display domains over small isolated areas.
During the second phase of this program, 47 garnet films were
deposited on GGG substrates using the R.F. plasma spraying method. In
general, domains were observed in most of the films. Defect density was
usually high and the surface of such films was uneven or faceted.
Five selected garnet films representative of the direct fabrication
method using the R.F. plasma were further characterized and the results
are summarized in Table 4. The spray process conditions used in growing
these epitaxial films from the flux are shown in Table 5. In all samples,
the variation in film thickness and the high defect density made magnetic
characterization diffirult using conventional techniques.
Sample films representative of the results obtained by the spraying
process were delivered to the Sponsor.
44
~-~-~~^>-~-*--*~^-~~-> ^^^.-^.... ^gujto
CQ <
Ö .ft >. n) o a w c
(II nj u j-; !-i
in tt) OH
OH
bi
Ü
B'6 r ° ^ u
'S
(U J2 *> >. ID
T3
in
6 i-H
ca
'S •4-) U
w 'S n
X
in
1=
. O
»J 3
u ^ n,
01 J3
E m w 2
M C n) M
o
Fil
m A
rea
Con
tain
ing
Do
mai
ns
o
r<1 -, >-H
CO *
■0° in
.1 0^
1 ^ id rj :l
=1 ^
M 5 BO g s,? DO g ö 5 rt 3 n) 3
h 0) M 4) H (U M «) 0 > O rH o >- 0 >
C
2 ? IT? ? 0) <U V (U h > h In
o
oo
NO oo
en
in vo
o 00
00
< < <
in
<
in (\J
12 <
c 01 fa JH
-x Ü i 1)
0)
XI 4-i
o -i X "O ffi ■^.
4-1 (L -4-J >- I«
>H IH m t) 111 R j-j +J 0) !> a 01
E Ä £ ^-< rt 0 nj
■n u TI r o « 1) •rH
>-H i-H
E XI XI X) JO XI X5 3 3 3 O
« w ta o
XI
in X -ö uO
n 3
fa io
n
Ga
Oe
tio
n f
ield
O
e
^ C ■ü <5
c " in cm o> m • rt nJ n a)
in S ^ 0)
£ u C > o
•H A
Sat
ui
Coer
S
trip
t3 X! in u X5 c 0) Ss ; OB 0) -1
45
,.^.^^J^^..—,.w.t.-.-^.^ ...J- ■...,- ^.^^-^^m***^ .—"—L--—iiiiiiriiiiiiniiii i .^^■^■^^-^IT..^. L.u..^^-.,-i..^l.^...-.-- ^ .:-.:
c n ig
E o» ■* m •<(> m
^ 01 «M IM _ M _
E MJ
1 X ♦J
0 "
Ü ?u ^ m T).
3 S' O^ O (Jv o^ o
h
>. o
iS E c . "] ° in r-
&H i ■ oj m tn »n -4
an
„•EL ) " ir' ti "> u ^
o o o o in rsi O o ON
o m m f\j —* FI
^ •"• ^
'S ■0 o
<
2Si o o o o o *h V »•S E"
■2 M U •
in m in GO ^
a s o a* CT^ o* o
s ^o<^ a, §
•r* O XI
V)
> Ü V 0
in |
Pow
d as F
l cf
h o o o o o
(0 C O
■f-t
to > (0 O
^3 1 o U
& fl 0) BJ 1H c a c u o S
M 60 Ul g s 00 bo
tfl a o w >»>.>, >> s Ui <ü ft Ä i5 ^S x X X tu tü Ü O O O o o 0 -a u ft
0)
u o o o o o W
ft IS O o o M fM rj IM m
2 D E > S S " a a 5 H S, 0 o 0 0
!? » bo do do k h (, b0 DO
ftg «<:<;< < <
h V ? „■ 0 «I ^ ft > ^ vO CM ^) *o in * U X O -^ ro en i- ft -1 s ^ (O ^ rn (M
•. o" ft 2 in so oo ^ in
« § •*-•«.* <M fvj
«
•x « o r^ c» o
£ E ffi ffi ie i2 S ,31 < < «< < <
46
■fcta. -- - .—J-. - —'■—-- ■ '-'-■"-^■•■■■'iihiiiilirilr-friri i...itaf j»,^, .,. ..»^i.u..-.. .y... ..».,.,: >jv -,,.„■■^,..^.^«1
— ^ : : i ! : 1 r -^ HI
5. 0 BULK CRYSTAL GROWTH
At the beginning of this program Gd3Ga5012 substrates of 15-18 mm
diameter were available commercially. Although the dislocation and inclu-
sion densities were acceptable the crystals contained an undesirable, highly
strained central core. The lattice parameter of the core was found to be o
0.001 - 0.002A larger than the surrounding region. The Czochralski growth
of Gd3Ga5012 developed rapidly during this program and now nearly perfect,
core-free crystals of up to 38 mm in diameter are available. Lattice para-
meter variations within the crystal of less than 0. 0005Ä can be maintained.
There appears to be no inherent reason why even larger crystals of the
same quality cannot be grown.
The conditions required to grow high quality, core-free gallium
garnet crystals have been adequately discussed in previous reports^1' 2)
and will not be detailed here. The optimum rotation and growth rates for
crystal growh were found to depend on the amount of furnace insulation,
thermal gradients within the furnace, and the relative diameter of the
crystal and crucible. Hence, the "optimum-conditions are peculiar to a
particular furnace arrangement and it is not possible to specify rotation
and pull, rates that can be used in general. These conditions must be
determined empirically for each furnace arrangement.
During the second period of this contract, high quality single crystal
of Gd3Ga5Oi2 were grown routinely to supply substrates for epitaxial films.
47
W™llf.HMm-,.m^a^1^>-tt|W||>M IMB a ' it mm --■■'--■■■ ii i..,.....J ■^...-. .;....,.,....- —. :.■..,....,.. ,.,.....,....., ■...„....,...,. j
iM-iiMKini-i-i-ui« M - in im<mmrmm'mi^immmmmm>i->">>'mi,< „ imtwm*mmmKm^mmmr*mmmmm*mmmmmmmmmmmmmim*'*''*H>mim
6. 0 CONCLUSIONS AND RECOMMENDATIONS
The technology of garnet bubble materials has advanced rapidly during
the last two years. This is due in great part to the development of LPE
growth of garnet films. Levinstein, et al. ,(11) discovered the remarkable
stability of supercooled solutions of garnet in PbO-B203 and first described
the simple "dipping" method of film growth. Since then, the technique has
been used extensively in the preparation of multicomponent, uniaxial garnet
films. This is the method used to prepare the high quality films supplied
the Sponsor during this project.
Most of the garnet LPE work has been done in PbO-based solvents.
However, Linares*22^ investigated other solvents including those based on
BaO. Recently, the BaO-based solvent has been the subject of an intensive
investigation by workers at Hewlett-Packard Laboratories. (23) This solvent
was also investigated in this work. (!) The BaO-based solvent exhibits
several desirable features including low volatility, n&n-corrosiveness and
low toxicity. However, the high viscosity of the melt at practical film
growth temperatures and the attendant difficulties of solvent drainage from
the film have been serious deterrents to its widespread use.
Epitaxial garnet films have been prepared by methods other than LPE,
including CVD, APS, sputtering and coprecipitation/annealing. The CVD and
APS processes have been studied as a part of this project. High quality films
can be prepared by CVD but it is extremely difficult to achieve run-to-run
reproducibility with the complex garnet films of interest to present day bubble
technology. Most workers in the field have now abandoned the CVD approach.
48
IM_ '^i^**^*^^****^^
■ iiii",pia ("n ■ .,.,-...1,..*- wmmimmmmim^mm™ <n'*'i> >'"'"• ~*~*~~^mimmBmmmmm*mammmmmmmmmw,mw*'mimm
In respect to APS. the major problem appears to be control of phase rela-
tionships. Epitaxial garnet films grown by APS are usually contaminated
with crystallites of orthoferrite. hematite, magnetoplumbite and/or other
solid phases. Low defect density films have not been grown and the
elimination of unwanted phases appears to be a formidable problem.
An understanding of the mechanics and kinetics of LPE film growth
has been developed^4' 25) and the reproducibility of the process has been
demonstrated. (26) Aii recent developments indicate that LPE "dipping"
growth of garnet films is a commercially viable process. Remarkable
progress has been made in the relatively short time the process has been
used to grow magnetic garnet films. It is reasonable to expect that
additional progress will be made in the future. It is believed that nearly
all workers in this field are now using the LPE "dipping" method and this
situation is expected to continue.
The growth of bulk, non-magnetic garnet crystals for substrates has
kept pace with the development of epitaxial film growth. Core-free single
crystals with dislocation densities less than 5 per cm^ can be readily grown
by the Czochralski technique. Such crystals, up to 38 mm in diameter, are
now available from commercial vendors.
A number of film compositions were delivered to the Sponsor during
this project. These films, tailored to support 5 - 7 ^im bubbles, represented
a variety of bubble materials. Included were films of primarily stress-
induced anisotropy, primarily growth induced anisotropy and films in which
49
b'"-^-" - — —^..T""
" ■ i" i i i " ■ i ■!■ 'I ■' i mmr^^mKimmtintmMaiiw JII 11 lima*|l«MIW|HPIW<mw«aM«*^^*V«^n«WWn«HOTMI^MVnnii^^^^^mn>.
both growth and stress-induced mechanisms were important. At least two
garnet systems. Eu-Yb-Y. and Sm-Y were identified as being of particular
practical interest. These garnets exhibit reasonable mobilities, satisfactory
quality factors and good temperature stability.
The "hard" bubble problem was not recognized when this project
was initiated and has not been a subject of this work. However, two
different, more or less auccossful solutions have been developed by others.
These are (1) the epitaxial deposition of a "capping" layer^27) and (2) ion-
implantation. (28)
Although a good understanding of garnet bubble films has been
obtained to date, additional work is needed. In particular, it is felt that
detailed study of the possible relationship between wall mobility and
turbulence and coercivity might be especially rewarding.
50
^»^^■..^.w:-.....:va.^^^^U^
..M.. "".M.". J...-..1I!. .i> \nwimmm*mi^*m****^mmmm^^mmm^memm~i~~m'~^**~~*'^^mmmmmmmiHmm*^^^mf
7 . 0 REFERENCES
1. J. W Moody. R. M. Sandfort, and R. W. Shaw, Final Report. Magnetic Bubble Materials, MRC-SL-341, Contract No. DAAH01-72-(>0490~ August 11, 1972
2. J. W. Moody, R. M. Sandfort, R. W. Shaw, and R. J. Janoweicki. Semi-Annual Technical Report. Magnetic Bubble Materials MRC-SL- 354, Contract No. DAAH01-72-C-1098, Feb. 11, 1973'.
3. R. W. Shaw, R. M. Sandfort, and J. W. Moody, Magnetic Bubble Materials: Initial Characterization Report. MRC-SL-326, Contract No. DAAH01-72-C-0490. March. 1972
4. R. W. Shaw. R. M. Sandfort; and J. W. Moody. Magnetic Bubble Material; Characterization Techniques Study Report. MRC-SL-339, Contract No— DAAH01-72-C-0490, July, 1972
5. D. C. Fowlisand J. A. Copeiand. "Rapid Method for Determining the Magnetization and Intrinsic Length of Magnetic Bubble Domain Materials", AIP Conference Proceedings. No. 5. 240 (1971).
6. R. W. Shaw. D. E. Hill. R. M. Sandfort, and J. W. Moody, "Determina- tion of Magnetic Bubble Film Parameters from Strip Domain Measure- ments". J. Appl. Phys.. 44, 2346 (1973).
7. A. A. Thiele, "Theory of the Static Stability of Cylindrical Domains in Umaxial Platelets", J. Appl. Phys., 41., 1139, (1970).
8. P. W. Shumate, Jr., D. H. Smith and B. F. Hagedorn, "The Tempera- ture Dependence of the Anisotropy and Coercivity in Epitaxial Films of Mixed Rare-Earth Iron Garnets", J. Appl. Phys.. 44. 449 (1973).
9. P. W. Shumate, Jr. , "Extension of the Analysis ^or an Optical Mangeto- meter to Include Cubic Anisotropy in Detail", J. Appl. Phys., 44, 3323 (1973). —
10. R. M. Josephs. "Characterization of the Magnetic Behavior of Bubble Domains". AIP Conference Proceedings. No. 10, 286 (1973).
11 H. J. Levxnstein, S. Licht, R. w. Landorf, and S. L. Blank, "Growth of High Quality Garnet Thin Films from Supercooled Melts" Aool Phys. Letters, 19_. 486 (1971). ' PH •
51
^^■v.^....,.....^.^.^^^ -,1,.,.,.,J...„:J..,..^.^,.,J.1...,.,.U........,.^-..:- -.■■. OM
'^»^»^VPWWlffffWIWIRWW'^S^tiPW'^SW^K^^
12. W. A. Bonner, J. E. Geusic, D. H. Smith, F. C. Rossol, L. G. Van Uitert, and G. P. Vella-Coleiro, "Characteristics of Temperature- Stable Eu-Based Garnet Films for Magnetic Bubble Application", J. Appl. Phys. 43^, 3226 (1972).
13. NASA Contract No. NAS1-11794
14. G. P. Vella-Coleiro, "Domain Wall Mobility in Epitaxial Garnet Films", AIP Conference Proceedings, No. 10, 425 (1973).
15. Jerry W. Moody, Roger W. Shaw, and Robert M. Sandfort, and R. L. Stermer, "Properties of GdyYs-yFeg-xGaxOjz Films Grown by Liquid Phase Epitaxy", Intermag 1973, Proc. , to be published.
16. B. E. Argyle and A. P. Malozemoff, "Experimental Study of Domain Wall Response to Sinusoidal and Pulsed Fields", AIP Conference Proceedings No. 5, 115 (1972).
17. J. C. Slonczewski, "Theory of Domain Wall Motion in Mangetic Films and Platelets, " J. Appl. Phys. 44, 1759 (1973).
18. E. Schloemann, "Light and Heavy Domain Walls in Bubble Films", AIP Conference Proceedings, No. 10, 478 (1973).
19. E. M. Gyorgy, M. D. Sturge, L. G. Van Uitert, E. J. Heil-Aer, and W. H. Grodkiewicz, "Growth Induced Anisotropy of Some Mixed Rare-earth Iron Garnets", J. Appl. Phys. 44, 438 (1973).
20. T. B. Reed, "Induction-Coupled Plasma Torch", J. Appl. Phys., 32. 821 (1961).
21. T. B. Reed, "Growth of Refractory Crystals Using the Induction Plasma Torch", J. Appl. Phys. 32 , 2534 (1961).
22. R. C. Linares, "Epitaxial Growth of Narrow Linewidth Yttrium Iron Garnet Films", J, Cryst. Growth, 34, 443 (1968).
23. see, for example, R. Hiskes and R. A. Burmeister, "Properties of Rare-Earth Iron Garnets Grown in BaO-Based and PbO-Based Solvents", AIP Conference Proceedings, No. 10, 304 (1973).
24. R. Ghez, and E. A. Giess, "Liquid Phase Epitaxial Growth Kinetics of Magnetic Garnet Films Grown by Isothermal Dipping with Axial Rotation", Mat. Res. Bull, 8, 31 (1973).
52
^.^..:,:..:, ......-■ "-.-^ ...■ ■ :....;.■ .-.^--...-..-.--,.. ^■u,-.!...-!.^-. .■,.■■■■... dsä^tt - -,...-..■.. ■■.....,. -■.,--^ ,.__^üÄ
■■-"■^fl.^ip.u^^^iiiV«'«,"ä»W^ i»?.i^.-^.."ir»'»v-«WTWL-■'■J"."Jj",!!»" *' i" .'l,*^M*,■^f■!^lW''-l-.■1^l■-■ w^^^ww^pwrw^^TFVT-Tjrsw^r^eww?^^^ ^ft»wj#-Jirr« i ■.^...jr,wi,1.wpji,-1M"pijv»ltiw'i|.«ii.) n JIWI^IJII uimi,. ^iMuipu^«iiJ!Ji..(iJiil.ium^sniqPi
25. S. L Blank, B. S. Hewitt. L. K. Shick, and J. W. Nielsen, "Kinetics of LPE Growth and Its Influence on Magnetic Properties", AIP Conference Proceedings, No. 10, 256 (1973).
26. B. S. Hewitt, R. D. Pierce, S. L. Blank and S. Knight. "Technique for Controlling the Properties of Magnetic Garnet Films", Intermag Conference, Washington D.C., 1973. to be published.
27. A. H. Bobeck. S. L. Blank, and H. J. Levinstein, "Multilayer Epitaxial Garnet Films for Magnetic Bubble Devices - Hard Bubble Suppression' Bell System Tech. J., 51, 1431 (1972).
28. R. Wolfe and J. C. North, "Suppression of Hard Bubbles in Magnetic Garnet Films by Ion Implantation", Bell System Tech. J.. 51. 1436(1972). —
53
■■-.■--'- «■-- ~- -' ■■-.■ ■■-■.■ ...■....^. ....-.,<—.. .„.. -.-■....,■.„-■.,„,..,.,
mmiii!r**r!rmzv?**Bm**y*W*W . *!!T^.r^*fl^-'l^l'^WWTO»v:w..i»V^
8.0 APPENDIX
8. 1 Growth Parameters for Eu0.sYbaa5Y2.35Fe3.8Ga1.2O12
Solution Composition:
Compound
PbO B203
Eu203 Yb203 Y203 Fe203
Ga203
TOTAL
Liquidus Temperature: Growth Temperature: Rotation Rate:
Wt
400.0 8.000 0.7374 0.2477 2.2237
25.2933 4.0491
440.6
976 C 970 C 200 rprn
Moles
1.792 0.1149 0.002095 0.0006285 0.009846 0.1584 0.02160
2.099
Mol %
85.37 5.474 0. 09980 0.02994 0.4691 7.546 1.029
100.0
8. 2 Growth Parameters for Sm0.36Y2.64Fe3.7Ga1.3O12
Solution Composition:
Compound
PbO B203 Sm203
Y203
Fe203
Ga203
TOTAL
Liquidus Temperature: Growth Temperature: Rotation Rate:
Wt
400.0 8.000 0.6293 2.6482
22.8199 4.0030
438. 1
977 C 970 C 2.00 rpm
Moles
1.7920 0.1149 0.001804 0.01173 0.1429 0.02136
2.085
Mol %
85.95 5.511 0.08652 0.5626 6.854 1.024
100.0
54
,-v--^......^ ■■■ - - ■- ■■"- ■ ■. ■ ....■.—.- ..-.,..... ,....^...,.. ,....„...^. ,,, ,„.., , -■■-■■.-:■. !•• IIH ■'r'i<l^■»■lW'-t::*a^*